美国南加州大学Remo Rohs团队近期取得了重要工作进展，他们研究开发了蛋白质-DNA 结合特异性的几何深度学习预测器（DeepPBS）。相关研究成果2024年8月5日在线发表于《自然—方法学》杂志上。

据介绍，预测蛋白质-DNA结合特异性是理解基因调控的一项具有挑战性但至关重要的任务。蛋白质-DNA复合物通常与选定的DNA靶位点结合，而蛋白质则以不同程度的结合特异性与多种DNA序列结合。这些信息不能在单个结构中直接访问。

为了获取这些信息，研究人员开发了结合特异性的深度预测器，DeepPBS。这是一种几何深度学习模型，旨在从蛋白质-DNA结构预测结合特异性。DeepPBS可以应用于实验或预测结构。可以提取界面残基的可解释蛋白质重原子重要性得分。当在蛋白质残基水平上聚合时，这些分数通过诱变实验进行验证。应用于靶向特定DNA序列的设计蛋白质，DeepPBS被证明可以预测实验测量的结合特异性。

总之，DeepPBS为机器学习辅助研究提供了基础，这些研究深化了人们对分子相互作用的理解，并指导了实验设计和合成生物学。

附：英文原文

Title: Geometric deep learning of protein–DNA binding specificity

Author: Mitra, Raktim, Li, Jinsen, Sagendorf, Jared M., Jiang, Yibei, Cohen, Ari S., Chiu, Tsu-Pei, Glasscock, Cameron J., Rohs, Remo

Issue&Volume: 2024-08-05

Abstract: Predicting protein–DNA binding specificity is a challenging yet essential task for understanding gene regulation. Protein–DNA complexes usually exhibit binding to a selected DNA target site, whereas a protein binds, with varying degrees of binding specificity, to a wide range of DNA sequences. This information is not directly accessible in a single structure. Here, to access this information, we present Deep Predictor of Binding Specificity (DeepPBS), a geometric deep-learning model designed to predict binding specificity from protein–DNA structure. DeepPBS can be applied to experimental or predicted structures. Interpretable protein heavy atom importance scores for interface residues can be extracted. When aggregated at the protein residue level, these scores are validated through mutagenesis experiments. Applied to designed proteins targeting specific DNA sequences, DeepPBS was demonstrated to predict experimentally measured binding specificity. DeepPBS offers a foundation for machine-aided studies that advance our understanding of molecular interactions and guide experimental designs and synthetic biology.

DOI: 10.1038/s41592-024-02372-w

Source: https://www.nature.com/articles/s41592-024-02372-w